Final Fantasy XIV’s Armoury system, yesterday and tomorrow

Final Fantasy XIV has always sold its class system on the idea that you use it more as an a la carte buffet. Abilities you learn in one class are useful in another class, and as a result, your Gladiator is a mix of several different abilities in a single package. The idea at launch was that mixing abilities and inherent mechanics would produce very different characters based on the needs of circumstance and your personal playstyle.

This is not what happened. Nor is it what happened following the large ability revamp, which actually wound up making cross-class skills less useful in many areas.

When the game relaunches, odds are good that the current system will be largely intact, at least at a conceptual level. (You have your class abilities and then a selection of abilities from other classes that you’ve learned, in other words.) And it’s my hope that on this third pass through the system, the development team gets things just right. But let’s take a look at the first two versions first.

I'm a Marauder using only Archer abilities!  The world's a crazy place.

 

Armoury at launch — in which class doesn’t matter

When the game launched, abilities were wide open. Every level, you gained another three points for equipping abilities, and each ability had a cost (usually about three points, although some were valued at two, and basic attacks were free.) You equipped whatever abilities you wanted so long as you didn’t run over your cost in abilities, and it was good.

Well, in theory it was good. In practice, it meant that for most classes, you would equip the same basic set of skills no matter what because they were just that good.

Let’s use the Gladiator’s Circle Slash as an example. Circle Slash and its later equivalent Circle Slash II were unbelievably good. They were cheap area damage for any melee class, a great way of thinning out groups or dealing with several enemies or just inflicting wide-scale damage. As a result, pretty much everyone playing a melee class needed to level Gladiator to pick those abilities up, and players then went on the bar of every class ever. Ditto skills like Moonrise, Second Wind, Bloodthirst…

Eric Heimberg once said, quite astutely, that if you let players pick any ability, they’ll inevitably just pick the good ones and ignore everything else. Out of a pool of 100 abilities, you’ll see 10 that actually get used. This was definitely the case at launch, when several classes benefited more by having ability bars completely eschewing the skills taught by the class itself. There was only a marginal benefit to using something true to your class; why in the world wouldn’t your Thaumaturge just equip the more useful Conjurer spells and work entirely as a Conjurer?

I’m glossing over the details here, but you get the idea. There was no reason to play as a specific class except for the handful of class-specific abilities. So this needed some reworking, hence the first major revision that changed the system.

Armoury until the end of the world — in which even fewer abilities matter

The revised system that Naoki Yoshida’s team put forth introduced several elements, including combos and a new system of gaining abilities. Abilities (and traits) from your class were automatically equipped. You also had a small number of discretionary slots for “extra” abilities, allowing you to customize your character without making your chosen class totally irrelevant. Each class had a distinct mechanical identity, and no longer did you have certain ubiquitous skills coupled with a handful of whatever for flavor.

No, instead you just had a small number of ubiquitous skills and a whole lot of cross-class cruft.

See, for what I imagine were very good reasons, combos did not work across classes. Even if you could equip two abilities that would normally combo (which was only occasionally possible), there was no actual combo option available. Coupled with this was an across-the-board removal of several weaponskills that were either too good, too specialized, or too universally desirable. This meant that we suddenly had only a few things to actually put on our bars with our spare room.

In one respect, this was good. My Thaumaturge would play very differently from my Conjurer. In another respect, this meant that I quickly ran out of useful cross-class skills because my Thaumaturge abilities were worlds better than the equivalent Conjurer skills and could combo off one another to boot. The problem wasn’t fixed; they’d just swung back in the other direction.

Let's not forget what these guys could do to the makeup.Armoury in a new world — how can this work out?

So one system is a bit too open, and the other is too restrictive. Clearly, the compromise lies somewhere in the middle.

For starters, I think abilities in general need to be more useful for various classes. Magic should be useful on physical characters for something other than a curative spell and a potential buff. I understand that no one wants my Archer to be casting spells all the time, but what about a knockback ability to keep enemies away from me when I’m fighting? Or a combo that allows me to remove the cast time from a spell as a Lancer, thereby making attack magic worth picking up?

Debuffs could serve quite a useful role here for character specialization, as well. If you take several debuff abilities on your bar, you could build as more of a control-style character rather than as someone who leads by raw damage.

Another major element we need is more choices. As it stands, several ability slots are taken up by what I think of as cross-class equivalencies, weaponskills that are exclusive to one class or another that perform the same function as another skill that can’t be equipped. Allowing cross-class combos would fix some of this trouble, but a better option would be to just eliminate workalikes by removing the need. Some of this seems to already be happening with the bit of Pugilist/Monk revisions the developers have hinted at.

In the end, more than anything, we need the Armoury System to live up to its potential. I think there are a number of ways to do that, but we need a few more options first, especially in the presence of Jobs and all of their limitations.